method: Alibaba-PAI V22018-09-29

Authors: Alibaba-PAI OCR team

Description: A seq2seq based image-to-text result by Alibaba-PAI OCR team.

method: Clova AI / Lens2018-09-13

Authors: Seolki Baek, Geonmo Gu, Jeongo Seo

Description: Our model is featured by CNN/RNN-based encoder and Hybrid CTC/Attention decoder. Moreover we proposed new text synthesis tools to make our model robust and high performance in the wild.

method: Baidu VIS2018-07-03

Authors: VIS-OCR

Description: A seq2seq based image-to-text result submitted by Baidu VIS-OCR, the former OCR team at Baidu IDL.

Ranking Table

Description Paper Source Code
DateMethodTotal Edit distance (case sensitive)Correctly Recognised Words (case sensitive)T.E.D. (case insensitive)C.R.W. (case insensitive)
2018-09-29Alibaba-PAI V2174.391983.92%129.369086.57%
2018-09-13Clova AI / Lens175.436783.00%132.422985.56%
2018-07-03Baidu VIS185.807882.85%150.852784.68%
2017-10-17Dahua OCR226.821982.76%179.257685.89%
2018-09-10Alibaba-PAI198.975081.32%160.942483.44%
2017-07-06Baidu IDL v3211.590980.02%171.151782.33%
2017-07-06HIK_OCR_v3191.247178.29%158.839980.12%
2017-07-05Baidu IDL v2219.194177.80%178.694580.26%
2017-07-03Tencent-DPPR Team & USTB-PRIR251.984076.31%185.357280.55%
2017-03-20HIK_OCR_v2300.339272.94%244.190375.69%
2017-06-29HKU-VisionLab258.586272.03%212.168574.19%
2017-03-11HIK_OCR318.873071.50%266.689474.05%
2016-06-23Baidu IDL351.525868.27%298.802870.92%
2018-12-19SAR437.164267.40%203.144678.82%
2016-01-29SRC-B-TextProcessingLab419.741262.11%367.122264.95%
2015-11-09Megvii-Image++508.832357.82%377.652163.99%
2018-04-25CNN+LSTM767.353534.28%623.189940.20%
2015-04-01MAPS1,128.007532.93%1,068.718433.90%
2015-04-01NESP1,164.496831.68%1,094.707132.98%
2015-04-02DSM1,178.614025.85%1,108.938127.97%

Ranking Graphic

Ranking Graphic